The Advantages of Floor Coupled Up-Firing Speakers

Pan said:

Without this ability our ancestors would have been tigerfood and we would not exist. :)

/Peter

that is correct :)
I am glad to see that the matter of "pinpoint imaging" in live music perfomances seems to be explained. At least all positions are explained :)

and what about loudspeaker-room interface? what about reflections, loudspeaker perception as distinct sound source due to crosstalk and reflections from side walls etc.

any more thoughts, experiences? :)
 
Yes!

Obviously the more diffuse and the more reflective the conditions are the harder it is to hear the location of a source with precision. This is true for live performances as well as reproduction.

Also the time and direction of the reflections is very important. Much work has been done on this of course.


Early reflections (around 5ms or so) has a strong influence on focus. Therefore if you are interested of hearing the true size and character of the recording venue, and also a realistic 3d soundstage with localisation of the recorded sources, these early reflections most be controled somehow in the listening room.

Search for ITD (initial time delay) and RFZ (reflection free zone).

I believe what drives many to a more diffuse reproduction is becasue a lack of understanding of how to extract the information form the recording. But sure.. it's much a question of preference.

I have tried various types of speakers and room treatment and even though I sometimes can apreciate a sound that consists of many reflections from the room surfaces IMO it gets tiring in the end becasue of the same signature is coloring every recording... "sameness".

I used the German Physics 360 degree driver for some time but later moved back to boxed speakers and then to dynamic dipoles for higher directivity.

If early reflections are controled and room resonances are dealt with.. and the "RT-60" is short enough than a two channel set up can open up a big hole into the recording venue which can be very realistic and impressive IMO.

I am looking into a surround set up right now though so I can have ambience all around me for more realism. 6-8 surrounds speakers.


/Peter
 
Pan said:


Early reflections (around 5ms or so) has a strong influence on focus. Therefore if you are interested of hearing the true size and character of the recording venue, and also a realistic 3d soundstage with localisation of the recorded sources, these early reflections most be controled somehow in the listening room.

/Peter

yes I am interested and this is why I try what I decribed above

Pan said:


I believe what drives many to a more diffuse reproduction is becasue a lack of understanding of how to extract the information form the recording. But sure.. it's much a question of preference.

/Peter

to make myself clear once again - I am not into "more diffuse reproduction"
quite te contrary - I am into precise and realistic imaging and soundstaging

Pan said:


I have tried various types of speakers and room treatment and even though I sometimes can apreciate a sound that consists of many reflections from the room surfaces IMO it gets tiring in the end becasue of the same signature is coloring every recording... "sameness".

/Peter

I have never observed "sameness" using the devices I desrcibed above. Rather I have observed "sameness" in standard stereo triangle set up in comparison

Pan said:


If early reflections are controled and room resonances are dealt with.. and the "RT-60" is short enough than a two channel set up can open up a big hole into the recording venue which can be very realistic and impressive IMO.

/Peter

this is what I have experienced with pseudoambiopole and what I aim to achieve in my new room with omnidorectional speakers

please take a look at my first post

- significant reflections off the adjacent wall would not reach the listener
- floor reflection would not reach the listener
- first reflections off the front wall and off the back wall would reach the listener 9.3 ms after the first wavefront
- first reflections off opposite walls would reach the listener 8.3 ms after the first wavefront
- first reflections off the ceiling would reach the listener 8.8 ms after the first wavefront (ear at height of 90 cm, ceiling at height of 300 cm)
It seems that apart from those earliest reflections <1 ms all other reflections are delayed more than 8 ms and caught by precedence(Hass) effect.
And all this in a relatively small listening room. The bigger the better.
Also I am considering using some wide band absorbers to reduce the reverb after first 50 ms as Moulton recommends.

this is how I try to "control somehow early reflections in the listening room"

what do You think?

best,
graaf
 
Pan said:
Graaf,

sorry for not paying enough attention to what you have written earlier in this thread. I was talking in general terms and did not adress you or your set up specifically. :)


/Peter

absolutely no offence taken at all :)

I asked You to take a look at my first post just because I would like to know Your opinion.
This is sort of alternative method of dealing with early reflections.
Alternative to moving the loudspeakers as far from the walls as possible and of treating the walls with absorbers/diffusors

best,
graaf
 
Hello Graaf!

I have been listening this way for a while, just with the drivers on ear-height. I am using the Ciare HX201, also not ideal and originally bought for an other kind of use. The problem is a 2kHz shout. There was a large 8" fullrange driver test in a German DIY magazine and all drivers with sufficient treble rise for compensating the ceiling loss have this shout. My recommendation is the Ciare HX160. According to an other magazine it is shout-free from 60° off-axis and has a similar rise as its bigger brother. The frequency response is very rugged in the treble, but the CSD is very fast without irregularities. The measured TSPs differ from those at the Ciare site:
Fs 59.56
Re 5.53
Rms 0.48
Qms 6.44
Qes 0.34
Qts 0.33
Cms 0.87
Mms 8.23
Bl 7.05
Vas 22.68
Le 0.05 (!)
Sd 137
In my simulations it works very well in horns, so hm´s Lure, which was designed for the DDD, could be an option if a BR happens to work not properly. I haven´t bought a pair of them or notched out the shout because I intend to get an EQ/ room correction system anyway for the OB-bass. In my former listening room the ceiling reflection had a delay of 4ms, which is the border between perceiving it as part of the source vs. early reflection. This worked VERY well. In my currrent listening room the delay is higher, which results in a desintegration of the spacial image. I will have to fix this with reflectors or wheels on the speakers to get it off the door while not listening (in case the room correction or multi-band delay can´t fix it).

regards,
Oliver
 
el`Ol said:
Hello Graaf!

The problem is a 2kHz shout. There was a large 8" fullrange driver test in a German DIY magazine and all drivers with sufficient treble rise for compensating the ceiling loss have this shout. My recommendation is the Ciare HX160. According to an other magazine it is shout-free from 60° off-axis and has a similar rise as its bigger brother. The frequency response is very rugged in the treble, but the CSD is very fast without irregularities.


thank You for that information, the Ciare driver looks very interesting

as to the "shout" - in my Fostex FE206e the main problem was sharp resonance at about 2.8 kHz
I treated the cone and it has moved up in frequency to about 4 kHz and is now also reduced in level. Definitely less offensive to the ear now

el`Ol said:

In my former listening room the ceiling reflection had a delay of 4ms, which is the border between perceiving it as part of the source vs. early reflection. This worked VERY well. In my currrent listening room the delay is higher, which results in a desintegration of the spacial image.

very interesting remarks as to beneficial effect of this earlier ceiling reflection - I have to try this

thank You very much! :)

best,
graaf
 
Hi Graaf, and others.

Is it not the case that as you walk closer to either LS, its centre beam plus the ceiling reflection of that beam seriously interfere, thus making an optimised listening area very small and narrow ?

Also what is actually being listened to, the front of cone radiation at possibly 50 degrees plus baffle edge radiation, for much more must go into room reflections than radiates directly ?

I'm not trying to pick holes or be funny, it is just that I had already tried a loudspeaker (B200 - also shouty) on the floor as suggested and it was so much inferior, especially with regard to the noted ceiling beam interferences.

Cheers ........ Graham.
 
Graham Maynard said:
Hi Graaf, and others.


Hello! :)


Is it not the case that as you walk closer to either LS, its centre beam plus the ceiling reflection of that beam seriously interfere, thus making an optimised listening area very small and narrow?


the question is what kind of LS?
but not only - also logically it is quite the contrary - when You move closer to the speaker, whatever kind, the ceiling reflection becomes more and more delayed and relatively lower in level in relation to the direct sound and thus less detrimental because more delayed and quiterer ceiling reflection less seriously interferes with the direct sound from LS
don’t You agree?
I don’t question that You can experience such a sound quality deterioration but I cannot find explanation for that


Also what is actually being listened to, the front of cone radiation at possibly 50 degrees plus baffle edge radiation, for much more must go into room reflections than radiates directly ?


this is a very VERY good question!
we should ask ourselves this question - what is sound that we are hearing?
in terms of amplitude vs frequency?
the correct psychoacoustical answer is that it is the energy of the direct sound plus energy of all reflections reaching our ears in the period of around 50 ms after the arrival of the first wavefront, after a sense of the direction of sound (a "sound source") has formed

and this is regardless of kind of loudspeakers because this is how our hearing works

It is a psychoacoustical fact that we never ever listen to something like "center beam of the loudspeaker" - with the exception of the case when we listen to it in an anechoic chamber, which happens not very often ;)


I'm not trying to pick holes or be funny, it is just that I had already tried a loudspeaker (B200 - also shouty) on the floor as suggested and it was so much inferior, especially with regard to the noted ceiling beam interferences.


inferior to what and in what terms? please give more details

perhaps the problem is not with interferences of early reflections but with later reverberation?

When You have a typical audiophile room i.e. a room that effectively absorbs sound in lateral plane (furniture, curtains, audiophile absorbers) and have quite (perhaps even the word "very" would be better) beaming, directional speaker like the Visaton B200 then directing the speaker up to the ceiling (which normally is totally reflecting) DRAMATICALLY raises the level of reverberation in the room
and that can make a BIG difference and can be perceived as "much inferior sound"

I’m quite sure about that

but it is all about being precise and not to generalize too easily from very specific cases :)

best,
graaf
 
c2cthomas said:
Oskar Heil Speakers uses speakers that fire up towards the ceiling and an Air Motion Transformer for the HF. http://www.unitedhomeproducts.com/id101.htm

In my Walsh 5 remake I am using the AMT above 5 KHz and an upside down 10 inch Pioneer for mids and LF up to 6 KHz. Below is a sketch of what the final version should look like - if I ever get around to making that version of the speaker. At present I'm considering other adventures which include OB. :cheers:

Shame to use such a powerful midtweet as a supertweeter. I'd consider cap-coupling it lower, you can mod the pioneer for less top end (cut the wizzer off and add some felt to surface of the diaphragm) and add an inductor if need be.

Oh, and if you haven't tried them yet: the heils are spectacular.
 
Graaf... how funny.. I didn't realise what driver you were using. Yesterday I looked after a 8" driver for a small guitar speaker to use with my PodXT and Tele/LP. I arrived at the FE206E as a hot candidate for this project. :)

I hope it will give a more open sound (due to size and dispersion) than the typical 12" celestion based Vox and Marshalls.

I can imagine your set up giving nice results due to the relative freedom from early reflections. Would like to try somehing similar some time. My current rooms are either to big or small for such a set up though I believe.


/Peter
 
Member
Joined 2006
Paid Member
Graff - you might find this design of interest http://gr-research.com/index.asp?PageAction=VIEWPROD&ProdID=113 plus there are a couple of other similar upward firing designs at that web site. Just food for thought as to what can be done with a DIY type of project. Combined with some room treatments some very good results should be obtainable. :D

badman - I use my AMT 1's for HF because that is what that particular design was specified for. Also the AMT-1's can have a bit of a nasty rise at around 3.5 KHz that would take a notch to filter out and I wanted to keep the XO simple (like me :xeye: ) These are the older models (1970's) and I would not want to get 'em lower than 1.5 KHz - but they are "sweet" - loved them since the '70 when I 1st listened to them. On my 2do list is buying some of the new replacement diaphragms and coming up with a new magnet arrangement that more refined than the old design. Sorry for the OT - :rolleyes:
 

Attachments

  • av1rs_med.jpg
    av1rs_med.jpg
    5.2 KB · Views: 2,271
Pan said:
I have done a lot of these tests in various acoustics and with various sources of sound. I write and record music, sometimes with a friend that also is a musician and use to produce music/radio jingles. We have recorded in rooms the size of a bedroom up to hangar sized halls. Building/arranging studios and listening rooms from 10 to 100 meter squared we have had great opportunity to test the room sound and localization of sound sources in close to anechoic to lively acoustics in the same room.


Without this ability our ancestors would have been tigerfood and we would not exist. :)


/Peter


That is the point - we don't need to localize the tiger to within 1m at 20m distance. Just a general direction. Our hearing is not designed for such precise localization (just compare design of bat's hearing apparatus and ours and differences will be rather obvious).
I go to live concerts on average twice a month (mostly classical, but some jazz - which is often amplified so it doesn't count).
I often try to listen with eyes closed (that is my "training") and without visual feedback I often find it impossible to identify which instrument is playing a particular line that I hear even if they are wide apart (e.g. woodwind section is usually spread over a whole back row). Try it next time - close your eyes and guess. You will be surprised.
 
Bratislav said:
I often find it impossible to identify which instrument is playing a particular line that I hear even if they are wide apart (e.g. woodwind section is usually spread over a whole back row). Try it next time - close your eyes and guess. You will be surprised.

Well as I mentioned I allready have made that many times, no surprises.

Now when it comes to woodwinds (I play woodwinds myself) those may be harder due to the relative non-transient nature. Also many sources of the same type tend to mix and become diffuse. What I allready wrote is still true and valid.

Obviously it becomes more diffuse and hard to pic the localization of individual instruments when the whole orchestra is going FFF.


/Peter
 
Pan said:


I can imagine your set up giving nice results due to the relative freedom from early reflections.

/Peter

Actually the method of dealing with early reflections I proposed above permits to achieve in a room of a given size results completely unachievable in a standard set up.

For instance:
Let’s assume stereo basis (distance between the speakers) of 3 m, which I consider as minimum for realistic reproduction, and a typical equilateral stereo triangle set up
In such a case the loudspeakers should be at least 1.4 m away from the nearest wall so that the first wall reflection would be delayed by 5 ms, which is considered bare minimum to avoid its detrimental effect.
It means quite a big room and dedicated exclusively for audio, because I somewhat cannot imagine a woman letting her partner to move a loudspeaker 1.4 m into the living room ;)

And Dr Geddes writes that actually the required delay is 10 ms to avoid the detrimental effects of early reflections
To achieve that result one really big room would be required, because the side walls would need to be about 7<8 m apart for that (assuming the absolute minimum stereo basis that is only 3 m wide!)
And also the front of the loudspeakers should be about 2 m away from the wall behind them.
Assuming that the listener is seated next to wall behind him (probably not very good idea but let’s assume) a room of 40 square meters dedicated exclusively for audio would be needed to satisfy the 10 ms delay of early reflection from side walls and back wall requirement. And what about ceiling reflection and the worst - floor reflection? We can possibly have a very high ceiling but floor reflection is inescapable with conventional speakers even in a very big listening room.
Listening room of 40 square meters dedicated exclusively for audio - quite impractical – isn’t it?
Not many music lovers could afford something like this. That is why people look for different solutions which aim at lowering the level of reflected sound like using reflection absorbers/diffusers but this devices also have no WAF at all and are not suitable for normal living room.
Another method of lowering the level of reflected sound is to control directivity of spekarer using waveguides as in Dr Geddes designs. Unfortunately the waveguides and matching woofers ought to be quite big to do the job. And resulting speakers are not very decor-friendly, quite visually obtrusive. And even in their case they ought to be moved away from walls and into the room! (please correct me if I am wrong)

In the set up I proposed only a room of about 25 square meters and a ceiling of about 3.5 m (or less if we move closer to the speakers) would be required to achieve the 10 ms early reflection’s delay. And absolutely no problem with floor reflection.

Moreover the speakers are to be positioned against the walls and only 20 cm above the floor – it means no WAF and home décor concerns at all, the speakers are practically invisible.
All that is required is:
- relatively free space around the speakers, that is no big/tall and/or sound reflecting furniture especially on the side walls (cupboards, tall bookcases etc., fortunately such a big and tall furniture is not characteristic for contemporary home decoration)
- listener’s seat to be located symmetrically with regards to speakers, which is a standard requirement for stereo reproduction and the sofa can be just moved to the right place for serious listening - no problem at all

best,
graaf
 
What can I say.. you cover it up nicely! :)


Some thoughts..

The floor reflection can be dealt with by using an array or with conventional speakers a table can be placed close to the listener to shadow the reflection (effective in the higher registers at least).

About the stereo triangel.. have you tried other angles? 22-23 degrees use to give best results for me with boxes and open baffles.


My solution?

I bought a house so I can go crazy with a dedicated room, big if needed.


/Peter
 
graaf, extent your idea even further.

Place LS in the corner, normally the worst possible position for a conventional speaker. Now, you get even more boost at low freq and your 2 walls and the floor can be assumed to be a BIG baffle, and at the same time the adjacent walls are as far away as possible.

Maybe a Visaton B200 stuck deep into the corner without enclosure or even baffle. Suspend the driver from the ceiling or mount it to a backbone of some sort. Stuff the walls next to the speaker and drive it with a transconductance amp.

Well, just an idea :D

JB